Dr. David Allison

speaker

936 appearances 1 recordings 1 series first heard Mar 2026 last heard 31 Mar

Dr. David Allison’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 1 in all, peaking in Mar 2026 with 1.

Appearances

newest first · ▶ plays the moment

The Dr. Gabrielle Lyon Show · GLP-1 Drugs & Bariatric Surgery Explained by a Top Obesity Researcher | Dr. David Allison · 31 Mar 2026 podcast

We say you can't have sex with everybody you want to have sex with under any circumstances.

Control yourself.

And we all think that's normal.

So why can't we think that, gee, you can't eat everything you want to eat and always feel satisfied?

Sometimes you'll feel a little unsatisfied.

Let's help you deal with that.

Sure.

This is particularly in the area of what specific techniques or programs have or have not been shown to work or how much they work.

My colleagues and I authored a paper.

Andrew Brown, my former mentoree, was the first author and I was the senior author on that, talking about ways to avoid exaggeration in the childhood obesity intervention world.

And it was a way of trying to guide editors and authors and readers of papers in saying these are many things that are done, again, intentionally or unintentionally, that lead to the perception that a treatment was more effective than it really was shown to be.

And we talk about things like inaccurate, incorrect statistical analyses of multiple types.

And I won't go through all of them here.

One particular I will mention is the analysis of cluster randomized trials.

So these are trials in which instead of randomly assigning each individual person

to a treatment, you assign clusters of people or groups, often their classrooms or clinics or towns or schools.

So you assign a whole school at once, and you might have 10 schools, let's say.

And it turns out that the statistics needed to analyze those are very different than the statistics needed to analyze when you individually randomize it.

And those cluster randomized trials, if analyzed correctly, have much lower statistical power than all other things being equal in individually randomized study.

And so people, the investigators, often don't like that because less statistical power means less chance of saying, eureka, I found something.

Showing 261–280 of 936 · page 14 of 47